michele-segata
Platoon at signalized intersection
Hello
I created some platoons at the signalized intersection.
And I saw that the platoons don't stop at red lights. only the leader vehicles stop. Could you show me how to fix this? Thanks
this is not possible without further ado. The car following model does not handle any traffic lights or regulations such as "right before left" or similar. You can work around this by instantiating the leader as a "DRIVER" and not as an "ACC". In case of a DRIVER, the Krauss model is used and traffic lights are considered. However, you will probably get further issues as the desired acceleration (as required by CACC) cannot be calculated and transmitted to the followers.
this is not possible without further ado. The car following model does not handle any traffic lights or regulations such as "right before left" or similar. You can work around this by instantiating the leader as a "DRIVER" and not as an "ACC". In case of a DRIVER, the Krauss model is used and traffic lights are considered. However, you will probably get further issues as the desired acceleration (as required by CACC) cannot be calculated and transmitted to the followers.
Thank you for replying to me. I changed the setSpeedMode to 31 instead of 0. traci.vehicle.setSpeedMode(vid, 31)
But only the leader car stopped abruptly, and the following cars kept going through the front car. Do you know a way to make the following cars also stop when the leader car stops? Thanks
And you can see, the simulator is lagged when the leader of platoon hits the red line
But only the leader car stopped abruptly, and the following cars kept going through the front car. Do you know a way to make the following cars also stop when the leader car stops? Thanks
Hey, Have you solved this problem?
But only the leader car stopped abruptly, and the following cars kept going through the front car. Do you know a way to make the following cars also stop when the leader car stops? Thanks
I guess the issue here is that, by changing the speed mode, SUMO forces the vehicle to stop, but the vehicle itself still computes an ACC desired acceleration which is shared with the others. As SUMO slows down the vehicles, the ACC algorithm will compute a positive acceleration, because it wants to speed it up to reach the desired speed. So the other vehicles believe the leader will accelerate, while it is actually slowing down. But to confirm this, you should do some debugging by printing the values exchanged between the vehicles.
thank you for replying to me. But how to print the value exchanged between vehicles?
I only know the position, speed and acceleration of vehicles over times
You first have to use the right controller. This would be either "CACC" or "Ploeg". ACC is not working since it is not suitable for platooning.
In case you use one of the two, the vehicle still relies on the 'controller acceleration', which cannot be provided by a human driven vehicle. This needs to be changed in the code, e.g. you use the real acceleration for the controller acceleration as well. This will very likely not keep the platoon string stable. Thus you somehow need to change the protocol for the beacon exchange or the properties of the controller. This is up to your scenario and your overall conditions.
You first have to use the right controller. This would be either "CACC" or "Ploeg". ACC is not working since it is not suitable for platooning.
In case you use one of the two, the vehicle still relies on the 'controller acceleration', which cannot be provided by a human driven vehicle. This needs to be changed in the code, e.g. you use the real acceleration for the controller acceleration as well. This will very likely not keep the platoon string stable. Thus you somehow need to change the protocol for the beacon exchange or the properties of the controller. This is up to your scenario and your overall conditions.
Thank you for replying to me. I modified Segata's model by adding traffic signals to the model. I would like to check the behavior of platoons when approaching the red signal. I understand that I have to change the protocol but I do not know how to change it. Could you provide examples about Plexe API with traffic signals so I can follow it? Thanks
My model here
Sorry for my late reply. What do you mean by
I would like to check the behavior of platoons when approaching the red signal. I understand that I have to change the protocol but I do not know how to change it.
It is not really clear to me what you want to do.
Hi All,
Even I'm facing the same issue, while vehicle reach the Junction/intersection... Vehicles are crossing the signal irrespective of phase of signal colour. I have tried to change the driver controller as well from IDM, ACC and CACC but none of them were working fine.
Could you please solve this or provide any solution that could help in solving this issue.
I guess the only way is to use the plexe::DRIVER controller in MSCFModel_CC for the leader, but then you still need to modify the car following model to be able to provide the followers with the required acceleration they need for control.
@TrinhTuanHung2021 @michele-segata @prasuka4 @Q-max007 hi all, I'm facing the same issue, vehicles do not stop at red lights at signalized intersections, have you solved this problem?
@TrinhTuanHung2021 @michele-segata @prasuka4 @Q-max007 hi all, I'm facing the same issue, vehicles do not stop at red lights at signalized intersections, have you solved this problem?
It related to the deceleration of vehicles in the platoon. If the leader stops suddenly, the following cars can not stop in time because the gap is too small. So I set the deceleration to be high to stop the following cars immediately.
The initial approach suggested by @michele-segata is to utilize the plexe:: DRIVER model. This will result in a scenario utilizing the actual acceleration rather than the desired acceleration. To resolve this issue, begin by setting the *.node[*].scenario.useControllerAcceleration value in the omnetpp.ini file to false.
This gives you a scenario where the actual acceleration is used instead of the desired one. As a consequence, this leads to a situation where the vehicles behind the leader in the platoon cannot respond "immediately" to the leader's actions, which could result in collisions. One potential solution is to increase the distance between the vehicles in the platoon.
Yet, by using the actual acceleration instead of the desired one, you cannot achieve the same 'performance' as in the ACC controlled scenario, but it will allow for successful stopping at red traffic lights.
The initial approach suggested by @michele-segata is to utilize the
plexe:: DRIVERmodel. This will result in a scenario utilizing the actual acceleration rather than the desired acceleration. To resolve this issue, begin by setting the*.node[*].scenario.useControllerAccelerationvalue in theomnetpp.inifile tofalse.This gives you a scenario where the actual acceleration is used instead of the desired one. As a consequence, this leads to a situation where the vehicles behind the leader in the platoon cannot respond "immediately" to the leader's actions, which could result in collisions. One potential solution is to increase the distance between the vehicles in the platoon.
Yet, by using the actual acceleration instead of the desired one, you cannot achieve the same 'performance' as in the ACC controlled scenario, but it will allow for successful stopping at red traffic lights.
much thanks for your reply, i will try
@TrinhTuanHung2021 @michele-segata @prasuka4 @Q-max007 hi all, I'm facing the same issue, vehicles do not stop at red lights at signalized intersections, have you solved this problem?
It related to the deceleration of vehicles in the platoon. If the leader stops suddenly, the following cars can not stop in time because the gap is too small. So I set the deceleration to be high to stop the following cars immediately.
much thanks, Can we exchange a wechat or something for better communication? i saw u on github so many times
Sorry - I mixed up the projects... The approach would work for the OMNeT implementation.
As I wrote in my answer above, you should use the Plexe::DRIVER controller for the leader and change the SUMO code so that you can fetch the acceleration of the human driver model via the API. Then you have to share such value with the followers in the platoon.